Material guide: hay absolute — the molecule that makes hay smell of hay became, once mouldy, the ancestor of warfarin

About a 5 minute read.

The short version

• Name: hay absolute, CAS 8031-00-3
• Odour type: hay; odour: sweet, hay, woody, tobacco, mossy, tonka
• Strength medium (rank 2), evaluate at 10% or less; substantivity 400 hours at 100% (measured neat)
• Suppliers: 34; 24 months, cool, dry, sealed
• Toxicity field: not determined
• The synonym field and the notes field name different plants

Two plants in one record

head_synonym: lolium perenne absolute — perennial ryegrass.

occurrence: rye grass. Consistent.

But the TSCA definition inside notes reads:

… hierochloe alpina, gramineae.

Hierochloe is sweetgrass, an entirely different genus from ryegrass.

One record, two plants. I hit exactly this in lavandin, whose TSCA definition named Convolvulus scoparius, a bindweed. The TSCA line in the notes field is this database's most frequently mis-pasted passage, and should be read as possibly belonging to another record.

(The two grasses are not unrelated in smell — sweetgrass is known for coumarin, and coumarin is the core of hay. Which is why this error cannot be caught by nose.)

The evaluation field reads like prose

odor_descriptions holds five entries, two of which do not sound like a database:

Meadow with wild grass and flowers in summer. Warm and sweet, honey and dry fig.

Herbaceous, honey, warm Virginia tobacco.

And another gives its position in a formula:

Hay has a soft, sweet, woody scent reminding tobacco and moss. HAY PAYS ABS is used in fern, aromatic and lavender fragrances.

It is a fougère filler. A neighbour of oakmoss, treemoss and coumarin — and the odour field names tonka outright.

And hay's smell has a hard history

Hay smells of hay largely because of coumarin. Standing grass carries little of it; coumarin is released once the grass is cut, dried and its cells break — another instance of the molecule not being what the plant left out for you.

But if hay does more than dry — if it moulds — things go a different way.

Duxbury and Poller open their 2001 review of oral anticoagulants in Clinical and Applied Thrombosis/Hemostasis:

Oral anticoagulation originated with the discovery of the harmful agent causing "sweet clover disease" in cattle in North America in the 1920s. The causative agent dicoumarol was isolated in Link's laboratory in 1940. A range of related compounds was then synthesized, the most popular of which proved to be warfarin.

Cattle ate mouldy sweet clover hay and did not stop bleeding.

The mechanism: sweet clover contains coumarin, mould converts it to 4-hydroxycoumarin, and two molecules of 4-hydroxycoumarin condense into dicoumarol — a vitamin K antagonist that prevents clotting factors being made.

Coumarin itself does not thin blood. The mould-modified version does. (That distinction matters, and I made it in the coumarin article too.)

The vitamin K half of the story is told by Norn and colleagues in 2014: Dam and Schønheyder in Copenhagen, feeding chicks on a synthetic diet during sterol metabolism experiments, found that a missing unknown factor caused internal bleeding — and that the bleeding was not reversed by vitamin C. In 1935 the factor was named vitamin K.

The two lines meet in the 1940s. One asks why cattle bleed; the other asks what makes blood clot. They are two faces of the same thing, and a relative of the molecule in between is still in the pharmacy.

This record's toxicity field is empty

toxicity_oral: not determined.

I measured this in article 100: the field looks 100% covered and is 12.3% once "Not determined" is stripped. This record is one of the other 87.7%.

So the history above is not toxicity data for this material. Hay absolute is not dicoumarol and it is not sweet clover — I tell that story because it explains what this smell is chemically, not because it says anything about this bottle.

How to use it

• Evaluate at 10% or less; the 400 hours was measured neat and can be compared directly with others.
• It is a fougère and lavender filler, in the same district as moss, tobacco and tonka.
• It is a natural extract (natural_synthetic reads natural), so treat batch variation as normal.
• Ask what plant you are buying. This record's two fields name two grasses.
• Coumarin is regulated in cosmetics (the EU lists it among the declarable allergens), and this is a coumarin-bearing natural — check the supplier specification and IFRA.
• 24 months, cool and dark.

What this doesn't establish

• I have not smelled it. This is fields, whole-database statistics, and two papers.
• Both papers are medical history, not fragrance research. I cite them for where dicoumarol and warfarin came from.
• "Hay's smell is largely coumarin" is prior knowledge; I have no GC data for this absolute and do not know its coumarin content.
• "Mould converts coumarin to 4-hydroxycoumarin, two of which condense into dicoumarol" is the standard account, not something I checked against the primary chemical literature.
• I cannot tell whether ryegrass or sweetgrass is this record's true source. Two fields say different things and I have no third source.
• I checked no IFRA or EU cosmetic regulation text on coumarin.